RGB LED Luminosity ?

[airman00]

I have the Sparkfun RGB LED - http://www.sparkfun.com/commerce/product_info.php?products_id=105
The LED has the following specs :

Luminosity (RGB):
Red Luminosity - 800 mcd
Green Luminosity - 4000 mcd
Blue Luminosity - 900 mcd

I assume Luminosity is brightness ?

So how do I modify the PWM for the Luminosity by each LED ? I assume the duty cycle for the Green would be less than the duty cycle of the Green luminosity, so that they are equal in brightness.

Can someone explain this whole luminosity thing regarding LEDs.

[ed1380]

luminosity is how much light it puts out at full power.
so at 50% duty cycle luminosity is halved

[airman00]

luminosity is how much light it puts out at full power.
so at 50% duty cycle luminosity is halved

So that means that 20% duty cycle of Green is equal to 100% duty cycle of Red? So I guess I should take that into account in the PWM, as I want the colors to blend .

[HDL_CinC_Dragon]

Correct

[paulstreats]

Dont forget that you can also run them through different value resistors in order to straighten out their brightnesses. It will make it easier to program....

[airman00]

Dont forget that you can also run them through different value resistors in order to straighten out their brightnesses. It will make it easier to program....

How can I figure out the proper resistor values?

[HDL_CinC_Dragon]

Red Luminosity - 800 mcd - 2v
Green Luminosity - 4000 mcd - 3.2v
Blue Luminosity - 900 mcd - 3.2v

The green color needs to drop voltage by 80% to equal the red brightness and the blue color needs to drop about 11%.
So whatever resistor you put for the red pin, just put an 80% more resistive one on the green pin and an 11% more resistive one on the blue pin.
If im not mistaken, that should keep them all balanced.

[paulstreats]

I would say to calculate the max value resistors using the normal resistor calculator, then im sure that you can relate the mcd values to the max resistance values and increase the value of the resistors appropriately.

so the red would want its max value resistor.
          green would want a resistor thats 500% of its max value (multiply the max resistance by 5)
          blue would want 112.5% of its max value

This is assuming that the mcd scales linearly.

so if the calculated resistor for the red led is 600 ohm's then use a 600 ohm. if the calculated value for the green is 550ohm then use a 2.75kohm (500% of the max value) and if blue was 550ohm then use a 619 ohm resistor

[Webbot]

Ariman00 - haven't we been here before in another post http://www.societyofrobots.com/robotforum/index.php?topic=5834.0where we've talked about all the maths and stuff?

So lets go back to basics. What are you trying to do?
How many brigtness levels do you need per LED?
How can you tell if 2 LEDs are the same brightness - by sight only?

[airman00]

Now lets assume that by doing what paulstreats suggested we end up with LEDs that give out the same luminosity ( measured in mcd).
Then another problem arises  (  : Green at 50% brightness and Red at 50% brightness might appear to be at different brightness levels because of color intensity to the human eye - Green might appear brighter than red, even at the same duty cycles. In the other thread it was suggested that I use another resistor, how can I calculate that resistor value?
Sensitivity graph:
 


I would judge brightness by sight - because of what I said above ( certain colors may appear brighter than others, even at same duty cycle). To tell if two things are the same brightness I might be able to use Roborealm , but I don't think I could use Roborealm because Roborealm does not have the same sensitivity to colors that the human eye has.

[Soeren]

Hi,

Use trial and error (i.e. a potmeter in this case) if you want to save time, you will not be able to calculate it precisely anyway.

If it's for a visual experience, "measure" with your eyes. If it has to be more correct, use a light meter that compensates for different wavelengths

[paulstreats]

remembering the rgd led's i have used, then you might be surprised to find that the mcd relation is correct and that the green does need around 5 x more resistance to provide a similar light.

Remembering that when I was getting a pure white light from them, the red and blue pots were almost off whilst the green pot was almost fully on.

I would recommend the trial and error approch tho...

ps, these were the rgb led's i have used:

http://uk.farnell.com/kingbright/kaa-3528embsgc/led-smd-plcc4-rgb/dp/8530203